deepinv.sampling#
This package contains various posterior sampling algorithms, including diffusion-based methods and MCMC methods. Please refer to the user guide for more details.
Diffusion models with Stochastic Differential Equations for Image Generation and Posterior Sampling#
User Guide: refer to Diffusion models with Stochastic Differential Equations for Image Generation and Posterior Sampling for more information.
Base class for Stochastic Differential Equation (SDE):min_num_steps |
|
Reverse-time Diffusion Stochastic Differential Equation defined by |
|
Variance-Exploding Stochastic Differential Equation (VE-SDE) |
|
Posterior distribution sampling for inverse problems using diffusion models by Reverse-time Stochastic Differential Equation (SDE). |
|
Preconditioned data fidelity term for noisy data |
|
The DPS data-fidelity term. |
|
Base class for solving Stochastic Differential Equations (SDEs) from |
|
Euler-Maruyama solver for SDEs. |
|
Heun solver for SDEs. |
|
A container for storing the output of an SDE solver, that behaves like a |
Custom diffusion posterior samplers#
User Guide: refer to Custom diffusion posterior samplers for more information.
Denoising Diffusion Restoration Models (DDRM). |
|
Diffusion PnP Image Restoration (DiffPIR). |
|
Diffusion Posterior Sampling (DPS). |
|
Turns a diffusion method into a Monte Carlo sampler |
Markov Chain Monte Carlo Langevin#
User Guide: refer to Markov Chain Monte Carlo for more information.
Base class for Monte Carlo sampling. |
|
Projected Plug-and-Play Unadjusted Langevin Algorithm. |
|
Plug-and-Play SKROCK algorithm. |